Team technical performance characteristics and their association with match outcome in elite netball

Objectives. There is little understanding of the optimal technical performance characteristics associated with winning matches in netball and how these might vary between competition levels and between each quarter of the game. This study aims to identify and compare team technical performance characteristics between elite domestic and international-level matches and among different quarters. Design. A retrospective cohort study was used. Methods. Netball data sets were analysed, containing quarter-level and overall match statistics from the professional league in Australia and international tournaments (n = 1,296 records, 648 matches). Permutation resampling tests were used to compare technical variables between domestic and international matches, and between winning and losing teams at both the match and (non-cumulative) quarter levels. Machine learning methods were used to model match outcome and to rank technical characteristics. Results. Differences in team technical characteristics were observed between domestic and international matches, relating to penalties, and score-related actions. There are a variety of team technical characteristics that vary between winning and losing performances, and their importance varies slightly between quarters within a match. A K-Nearest Neighbour classifier explained match outcome with a high accuracy which demonstrates that there are enough data available in elite netball to apply machine learning methods for more complex analyses in the future. Conclusions. Team technical performance varies between winning and losing performances, at elite domestic and international levels. Minimising undesirable technical actions such as penalties and errors appears to be the most important difference between winning and losing teams.